Table of Contents

Built‑in Tools

AIVAX provides a list of built‑in tools that you can enable in your model. These tools can be used together with the server‑side functions.

Some functions have a cost. This cost is applied to models used by AIVAX and those you provide through BYOK (bring‑your‑own‑key), so it’s important to add balance if you intend to use these tools.

Note that each model decides which function to call and its parameters. Not all models may obey the call rules.

This function enables web search in your model. With it, the model can look up specific or real‑time information, such as weather data, news, game results, etc.

Web search is performed by several providers, chosen according to network availability and latency. AIVAX uses a mix of providers to carry out web searches.

AIVAX offers two configurable search types via its dashboard:

  • Full: the performed search is complete, inserting the entire content of each result found into the conversation context.
  • Summarized: the performed search is summarized, inserting into the conversation context a summary generated by AI from the provider itself.

The cost of both modes is $5 per 1,000 searches performed. The Full mode may consume more input tokens in the conversation, but can provide more precise results.

Note

Important: the Full search is not always available.

This function allows the model to have a web browser and perform automation, such as searches, navigating sites, filling forms, and clicking buttons.

There is currently no parameterization for this function.

The cost of this function is $2 per hour of automation, plus a fixed amount of $11.10 per 1,000 calls of this function.

Code Execution

This function allows the model to execute JavaScript code and inspect the execution result. With it, the model can evaluate, via algorithms, results of mathematical expressions and other situations that are better represented through code.

The code runs in a protected environment with very few functions available. The model will not be able to access I/O, the internet, or import scripts through this tool.

This function has no cost.

URL Context

This function allows the model to access external content at URLs and links provided by the user. With this function, the model can fetch links and evaluate their content.

Note that some destinations may identify the access as a bot and block it, as this function is not a crawler but a simple GET request to the target.

The model can access up to 5 links at once. Only the first 10 MB of each link are read. When the link content is retrieved, the system checks the returned content and handles it according to its type:

  • HTML content is rendered: HTML tags, scripts, CSS, and “noise” are removed from the access result, leaving only the plain text of the link.
  • Other textual content: the content is read directly and no transformation is performed.
  • Non‑textual content: when the link responds with non‑textual content and the response indicates a filename (either via the path or the Content-Disposition header), the system attempts to convert the downloaded file to a textual version.

This function has no cost.

Memory

This function allows the model to store relevant content for use across multiple conversations.

Currently, this function is only available when used in chat clients and when the session is identified by a tag.

Through the session tag, the model stores a relevant piece of conversation data, such as name preferences, reminders, or actions the assistant should perform.

The instruction for this memory tells the model not to save sensitive or personal data; however, it is not guaranteed that the model will always follow this rule.

Data are stored for one year in AIVAX’s databases and can be deleted at any time by the platform. For every conversation via a chat client, this data are retrieved and attached to the conversation.

This function has no cost.

Image Generation

This function allows the model to create AI‑generated images.

AI‑generated images are attached to the conversation context, but are not directly visible to the assistant.

Currently, there are four categories of generated images:

Image quality

  • Low quality: generates images quickly with low cost, but may produce many artifacts, such as extra fingers, distorted eyes, misplaced arms.
  • Medium quality: a balance between good quality and speed, though artifacts may still appear.
  • High quality: higher image quality and lower chance of artifacts.
  • Ultra‑high quality: the highest quality in image generation. After generation, an upscaling model refines the created image.

This function has a cost. The cost varies with the processing time of each image, which is affected by compute consumption, processing queue size, models used, etc. Image generation occurs on an external provider, so the cost can change based on many factors.

The example image above shows a price forecast for each image quality.

You can also enable the generation of explicit and adult images. When this feature is enabled, the model is allowed to generate adult material. For this to happen, the model must also “agree” to generate such content. Some models have a lower safety filter than others. For example, Gemini models have the lowest safety filter, making them a viable option for role‑play and generation of this type of material.

You are always responsible for the material you generate and the generated material must comply with our terms of service.

Generated images are stored on AIVAX’s servers for a few months before being permanently removed.

X Post Search

This function allows the model to search for posts on X (formerly Twitter).

It is a direct alternative to web_search, as it can be used to look for up‑to‑date real‑time information, such as news, data, game results, etc. This tool returns much more recent results than conventional web search.

It is not recommended to use both functions together because they have the same purpose.

Currently, the latest 20 posts on a given topic are inserted into the conversation context, including link and author.

At the moment, it is not possible to access posts from specific profiles.

The cost of this function is $5 per 1,000 searches performed.

Document Generation

This function allows the model to create PDFs from HTML text.

The created files are hosted on AIVAX’s servers and made available by the assistant.

The content is hosted for a few months before being permanently deleted.

This function has no cost.

Web Page Generation

This function allows the model to host HTML pages on AIVAX’s servers.

This enables the model to host reports, landing pages, and other HTML infographics.

The content is hosted for a few months before being permanently deleted.

This function has no cost.

Edit this page